PSI - Issue 39

R. Yarullin et al. / Procedia Structural Integrity 39 (2022) 364–378 Author name / Structural Integrity Procedia 00 (2021) 000–000

369

6

3. Numerical study The numerical study in this paper was carried out by means of the commercial FRacture ANalysis Code 3D, FRANC3D. The latter is based on FEM approach and designed to simulate 3D crack growth in engineering structures where the component geometry, local loading conditions, and the evolutionary crack geometry can be arbitrarily complex. 3.1. FEM models The cracked cylindrical specimen was modeled by ABAQUS and FRANC3D software using different types of elements (C3D8, C3D10, C3D15, C3D20). In order to guarantee a correct reproduction of the experimental loading conditions, the machinery clamps were replicated by using a kinematic coupling constraint between the specimen’s extremal lateral surfaces and two Reference Points (RPs) located on the top and the bottom surfaces. A kinematic coupling constrains the motion of the coupling nodes to the rigid body motion of the reference node. Therefore, the Boundary Conditions (BCs) and loads were applied on the RPs. Fig. 5 depicts the cracked FE model undergoing a pure torsional load at the last propagation increment. Here, the clamped and loaded surfaces, the domain and the top RP are addressed. Furthermore, a close-up of the surface crack tip is also provided.

Fig. 5. FEM model for cracked specimen undergoing pure torsion loading condition.

The pure tension tests were replicated by constraining all degrees of freedom of the bottom RP whereas the top RP is allowed to move only along axial direction. In addition, an axial force acting on the top RP was used to simulate the tension loading condition. A tensile load of magnitude 35 kN and a pure tensile load of 80 kN were respectively applied to D16T and B95AT Al-alloy specimens. The tension-torsion tests were modeled analogously by clamping the bottom RP and by constraining the top RP to slide only in the axial direction. Moreover, the tensile force as well as the torque around the specimen axis have been applied on the top RP, with a magnitude of 40 kN and 250 Nm, respectively. Both Al-alloys were tested in tension torsion with the same load levels. The pure torsion tests were modeled by clamping the bottom RP and the torque around the specimen axis was applied on the top RP, with a magnitude of 450 Nm and 250 Nm for D16T and B95AT, respectively. The geometry of the numerical initial crack replicates the specimen notch: it is modeled as an ellipse, cutting the specimen orthogonally to its axis at middle height. The elliptical notch dimensions, a and c , were set to sizes equal to 3 mm and 8 mm, respectively. Fig. 6 shows the final crack shape, in the case of pure tensile load, with dimensions definition: a, b, c .